解决方案查找器 技術支援

Wrong calculus of primary inductance ?

Posted by: cabedikes on
Hello!. I'm designing a SMPS with TOPSwitch-GX/HX devices. I have installed PI Expert Suite on my PC, I have started to work, and I have seen the following results : 1. When I have used a TOPSwitch-GX device and PIXIs software,I have obtained LP = 2587 uHenries = 2.587 mHenries When I have used a TOPSwitch-GX device and PIExpert software, I have obtained LP = 2403 uHenries = 2.403 mHenries Why these results are so different? I think that calculus of primary inductance on PIExpert with a TOPSwitch-GX is wrong because do not use 'fs(min)', but 'fs', when applying LP = 10^6 x Po / (IP^2 x Kp x (1-Kp/2) x fs(min)) x (Z x (1-n) + n)/n. Am I right? 2. And when I have used a TOPSwitch-HX device, I have obtained LP = 2315 uHenries (PIXIs and PIExpert software) This result do not coincide with any of the others two, why? Thank you very much for your assistance! Used datas (all cases) : VACMIN=112Vac fL=50Hz VO=10V PO=30W n=0.8 Z=0.5 tC=2.40ms CIN=56uF KI=1 Frequency=H VOR=100V VDC=10V KP=0.4 LP Tolerance=0 Obtained results (all cases) : DMAX=0.47 IAVG=0.31A IP=0.81A IR=0.33A IRMS=0.45A

回應

Submitted by PI-Spock on 11/03/2008

The value of primary inductance (LP) is a function of the "degree of continuousness" of the design (i.e. the parameter KP)
One difference in both the designs will be the value of KP which is a user input in PIXls. In PI expert you get Lp = 2.4 mH and in PIXls you get 2.5 mH. This tells me that you are probably using a smaller value of KP in PIXls than in PIExpert.
This is a very common question mainly because in PI expert KP is automatically optimized for each individual design by the software whereas in PIXls the value of KP is a user input. Most users do not change this value which leads to the difference in value of primary inductance.
Hope that clears things up.

Submitted by cabedikes on 11/04/2008
I understand that if I use different values of Kp, I will get different values of LP. I send you my design files (copied on two Word files), you could verify that I use the same value for Kp in PIXIs and in PIExpert. Why I don't get the same value of primary inductance (LP)? Thank you very much for your assistance

PI Expert uses a slightly different switching frequency as compared to PIXls. PIXls uses jitter frequency in its computation (hence switching frequency is smaller in PIXls as compared to PI Expert) which results in the slightly higher primary inductance. The frequency jitter component can be ignored for TOPswitch based applications because you typically always have some room to the current limit of the part.